Detection of a point mutation, a so-called single nucleotide polymorphism (SNP), is employed widely as a method of analyzing, at the gene level, for example, the causes of all types of diseases and the individual differences in disease liability (susceptibility to diseases) and in drug action.
Polymorphism detection methods that generally are practiced include (1) direct sequencing in which, in connection with the target DNA of a sample, a region to be detected is amplified by PCR (polymerase chain reaction) and the entire sequence of the amplification product thereof is analyzed; (2) PCR-RFLP (restriction fragment length polymorphism) in which PCR is performed in the same manner as in item (1) above, the amplification product is treated with restriction enzymes, and the change in restriction fragment length due to the polymorphisms is subjected to typing by Southern hybridization; and like methods.
However, the method of item (1) above requires, for example, sequencing after PCR and then electrophoresis or the like by a sequencer. The detection is thus very troublesome and costly. Moreover, the resulting amplification product needs to be subjected to treatment after PCR and may be contaminated during such treatment. The method of item (2) above also requires treatment of the resulting amplification product with a variety of restriction enzymes for analysis after PCR, thereby being troublesome. Moreover, the treatment of the resulting amplification product with restriction enzymes needs to be performed after the amplification product is transferred. Therefore, it may be possible that the amplification product obtained in a first reaction is scattered and may find another way into a second reaction that is performed separately. These problems cause another problem in that it is difficult with the methods of items (1) and (2) to automate the detection of point mutation.
Addressing these problems, a Tm (melting temperature) analysis recently has been attracting attention as a method for detecting an SNP. In this method, first, using a probe that is complementary to a region containing the SNP to be detected, a hybrid (double-strand nucleic acid) between a sample nucleic acid and the probe is formed. The hybridization product then is subjected to heat treatment, and the dissociation (melting) of the hybrid into a single-strand nucleic acid in response to a temperature increase is detected by measuring a signal such as absorbance. This is a method for determining an SNP by obtaining a Tm value based on the result of detection. The higher the homology of a hybridization product, the higher the Tm value, and the lower the homology, the lower the Tm value. Therefore, when the polymorphism of a detection target site is X or Y, a Tm value (reference value for evaluation) is obtained in advance in connection with a product of hybridization between a nucleic acid containing the desired polymorphism (for example, Y) and a probe that is 100% complementary thereto. Then, the Tm value of the sample nucleic acid and the probe is measured (measured value). When this measured value is identical to the reference value for evaluation, the sample nucleic acid and the probe perfectly match. That is, the detection target site of the sample nucleic acid can be determined as being of the desired polymorphism (Y). In contrast, when the measured value is lower than the reference value for evaluation, the sample nucleic acid and the probe mismatch. That is, the detection target site of the sample nucleic acid can be determined as having the other polymorphism (X). With such a method, an SNP can be detected, for example, only by subjecting a PCR reaction solution to which such a probe is added to thermal treatment and performing a signal measurement, and it is thus possible to automate a detection device.
However, such detection methods using Tm analysis have problems as follows. Generally, a gene polymorphism is present in the form of a homozygote (for example, X/X or Y/Y) or a heterozygote (for example, X/Y). It is important in the detection of a polymorphism to distinguish between a homozygote (X/X or Y/Y) and a heterozygote (X/Y) and, in the case of a homozygote, to distinguish between an X/X homozygote and a Y/Y homozygote. In the case of a heterozygote, gene polymorphism X and gene polymorphism Y are included, and the difference between these polymorphisms is merely a point mutation, i.e., a difference of one base. Accordingly, the following phenomenon occurs: a probe that fully hybridizes with a sequence that contains one polymorphism (for example, Y) (perfect-match) also hybridizes with a sequence that contains the other polymorphism (X) (single-base mismatch). In a case like this, there is a problem in that, as shown when a melting curve that indicates a relationship between signal intensity and temperature is drawn up based on a Tm analysis, it is difficult to detect a peak on the low-temperature side that indicates a mismatch sequence due to the presence of a peak on the high-temperature side that indicates a perfect-match sequence. That is, even when there is a mismatch sequence in a sample, the presence of a perfect-match sequence makes it difficult to distinguish the mismatch sequence, and thus it may be possible that the detection sensitivity is impaired. Also with respect to homozygotes, the difference between a homozygote (X/X) of a polymorphism X and a homozygote (Y/Y) of a polymorphism Y is likewise due to one base. Therefore, as described above, when it is difficult to distinguish between the peak of a perfect-match and the peak of a mismatch, it is consequently difficult to distinguish between a peak indicating the former (X/X) and a peak indicating the latter (Y/Y). That is, it may be possible to determine whether a sample is a homozygote, but it is likely to be difficult to determine the type of polymorphism.
Recently, antibody drugs that take advantage of the human immune function have been attracting attention in the field of pharmaceuticals. Examples of such antibody drugs include trade-name rituxan (generic name: rituximab), which is a therapeutic agent for malignant lymphoma, trade-name herceptin (generic name: trastuzumab), which is a therapeutic agent for breast cancer, and like agents. However, the strength of human immunity varies from person to person, and this is considered to have an influence on the pharmaceutical effects of such antibody drugs. Gene mutation (for example, SNP) is reported as a factor that has an influence on the pharmaceutical effect of such drugs. In particular, a polymorphism of the FCGR3A gene (Non-Patent Document 1), which is a gene that is involved in immunity, has been reported. This gene codes for FcγRIIIa, a type of fragment C receptor (FcR) of IgG. Hence, when such an antibody drug is used in medical treatment, it is considered useful, for example, to detect polymorphisms (SNPs) in such a gene and then to determine the course of medical treatment, e.g., the dosage of the antibody drug and the change of a therapeutic agent, taking into consideration of the results of the detection.    Non-Patent Document 1: Buillaume Cartron et al., BLOOD, 1 Feb., 2002, Volume 99, Number 3